We need a multi-national, publicly funded research organization akin to CERN/within CERN, whose whole purpose is to develop a state-of-the-art browser that's not Chromium-based. Make follow our lead, rather than us having to follow Google.

If the Web could be developed using public money, why not a modern browser? Public funding would remove the Mozilla problem of them having to depend on Google.

With the amount of money governments waste annually, we could fund this AND Mozilla.

There could be incentive problems here as well, of course, like governments threatening to withdraw funding in case a certain backdoor isn't included, or if it blocks ads too aggressively and some corporate-funded 'representative' starts receiving pushback from the industry etc, but which is why it would need to:

- Be funded by a wider variety of states than the Five/Nine Eyes members.

- Developed entirely in the open, each important change reviewed by a committee of experts from the public.

Show thread

@MatejLach But how would you unseat Chrome at this point? Google have the incumbent advantage and the platform advantage. Technical excellence is only part of the story.

@cbowdon That's definitely going to be a challenge, but did some smart marketing by having ads IRL, like in trains and such, even in smaller countries if the % of connected users was high enough.

Since it would be publicly funded, you could also install it on computers in publicly-funded educational institutions. A lot of software spreads by children installing it for their parents. If students are using it at school, they're likely to install it at home.

@MatejLach Ooh that last one is a good one. That’s what MS/Apple/Google are trying after all. You wouldn’t necessarily need CERN-like levels of funding to achieve it.

@cbowdon @MatejLach
but wouldn't you need CERN-like levels of fuding to develop a browser that keeps up with the moving target of shitty WHATWG standards?

@MatejLach @Wolf480pl @cbowdon

Also why build a state of the art shit whose shape has beed already defined by #Google instead of building something new and better?
Something following a totally different vision?

I think to pull regular users in, we'd have to start with today's web. But once we have sway in the committees, you can begin to redefine what state of the art web should look like.

@Wolf480pl @cbowdon

@MatejLach @Wolf480pl @cbowdon

I think we need a #CERN of #Informatics, but it should start from a simple vision and build what it takes to get there from the ground up.

I have a vision to propose: all people should be able to read, understand and modify each software they use or feed with their data.

Modern Web is not going to survive such vision, so building a browser is wasting money imho.

@Shamar@mastodon social It won't work. Just take some time to, say, explain recursion or graph algorithms, image compression or even cryptography math to a totally untrained user. We will never get to a point of end users to read or understand their software. IMHO, trying to do so is a waste of time that could better be spent on building more ethical solutions that just work for this crowd.
@MatejLach @Wolf480pl @cbowdon

@z428 @MatejLach @Wolf480pl @cbowdon My position is that they should be *able* to (perhaps with a little training), but not obligated to.

@alcinnz @z428 @MatejLach @Wolf480pl @cbowdon

Guys, that's just because we are at the hieroglyphs of #Informatics.

If it's difficult to explain it's because it's primitive. Let's invent the right alphabet and every kid will be able to learn programming at the primary school.

@Shamar We're at a point where some adults have issues understanding higher math, some even have real issues learning to master natural language to understand complex texts or express themselves. And we actually did invent an alphabet to help these folks: Icons. Symbols. Easy interactions. So far this works well. Will we be able to do meaningful programming on that level?
@alcinnz @MatejLach @Wolf480pl @cbowdon

@z428 @Shamar @alcinnz @MatejLach @Wolf480pl @cbowdon in simpler systems, the meaning of "meaningful programming" might be a lot different than it is in bloated corporate software. just want to get that noted.

@grainloom @z428 @alcinnz @MatejLach @Wolf480pl @cbowdon

We need to be like Moses.

We can all see how badly broken is current IT.
We can all see how much power we have (which ultimately is much much more we are fooled to think).
We call all see how hard corporations try to lock us in, layer over layer.

Can we think the promised land?
Just like ancient scribes couldn't think of a phonetic alphabet.

@grainloom @z428 @alcinnz @MatejLach @Wolf480pl @cbowdon

But we can try new roads.

We can experiment.

We can teach kids that they can reinvent the future in a different way.

Not just with our lessons but with our code and our example.

It IS possible.
Yes there's a lot of complexity to subdue, we still lack fundamental tools like Egyptians lacked the number zero.
But we need #hope to look for them! ;-)

@Shamar I think we very often fall victim to oversimplification because we have totally lost sight of how incredibly much specialized we already are - and how extremely basic and "trivial" some of the issues users are struggling with actually are. Google, Apple, ... are successful because they do better here, no matter why they do that.
@grainloom @alcinnz @MatejLach @Wolf480pl @cbowdon

@z428 @grainloom @alcinnz @MatejLach @Wolf480pl @cbowdon

Im 6 hour I teached to 22 yo kids what is defined at tesio.it/documents/vademecum.t plus basic networking (IP packets, IP addresses, DHCP, DNS and routing).

We did a simulation of packet routing with paper packets and they understood MitM and DNS poisoning by themselves.

The teacher proposed to add an our to explain one time pad encryption.


@Shamar And now, provide those kids with, say, a batch of hardware and the most simple fully featured implementation of something like e-mail. Do you think they will have a chance to understand what happens, let alone fix it? If that was possible, most programmers apparently are pretty dumb, looking at how much time is spent on fixing ...
@grainloom @alcinnz @MatejLach @Wolf480pl @cbowdon

@z428 @Shamar @grainloom @alcinnz @Wolf480pl @cbowdon

My take on that is that most of the required info is already out there, but I am all for simplifying it.

I don't think that would lead to some massive influx of programmers, because some people just have different passions like painting, music and such and some just want to watch TV.

There's a pretty large artist community on the Fediverse, don't think they're much interested in the tech side and that's honestly fine.

@MatejLach @z428 @Shamar @alcinnz @Wolf480pl @cbowdon idk, you need to know a lot of "engineeringy" stuff to get things done with digital art tools
you need to use logical thinking for setting up complex things in Blender or Krita
we are already forced to learn a large subset of MS Office, why couldn't we learn UNIX(or hopefully Plan 9) instead?


@grainloom @z428 @Shamar @alcinnz @Wolf480pl @cbowdon

Well, am not saying we shouldn't But many people learn a specific tool to achieve a specific tasks. Things like operating systems, programming and such are such open-ended things that unless you're creatively interested in them, there doesn't seem to be as much of a point in doing so.

@MatejLach @grainloom @z428 @Shamar @alcinnz @cbowdon

I think the primary reason why we'd want to teach everyone programming isn't so that they come up with new ways of implementing some part of an operating system, but so that they can make minor adjustments to the software they use on a daily basis, without being dependent on other programmers' willingness to implement such a feature or fix a particular bug.

@MatejLach @grainloom @z428 @Shamar @alcinnz @cbowdon

It'd be awesome to get to the point where "fork it if you disagree" is a viable option for everyone.

@Wolf480pl Yes, but way more than that I would hope people finally got away from that high frequency forking and "my way or the highway" kind of community to some sort of actual corporation again, willing to iron out even personal differences and come to a consensus. But that's possibly another thing.
@MatejLach @grainloom @Shamar @alcinnz @cbowdon

@Wolf480pl Yes, but way more than that I would hope people finally got away from that high frequency forking and "my way or the highway" kind of community to some sort of actual cooperation again, willing to iron out even personal differences and come to a consensus. But that's possibly another thing.
@MatejLach @grainloom @Shamar @alcinnz @cbowdon

@z428 @MatejLach @grainloom @Shamar @alcinnz @cbowdon
Maybe it's just me not looking in the right place, but I don't see too many forks recently...

Anyway, IMO it's better for a program to have a coherent vision behind its design, rather than have design-by-committee. And how do we check which vision is better, if not by forking?

@Wolf480pl @z428 @MatejLach @grainloom @Shamar @alcinnz @cbowdon honestly, a BDFL 'vision' model isn't much better than a traditional committee model

@a_breakin_glass @Wolf480pl @z428 @MatejLach @grainloom @Shamar @cbowdon Either approach can go too far.

Personally I lean towards a BDFL 'vision' model, but not if they don't listen to some committee and fold that into their vision.

@alcinnz @a_breakin_glass @z428 @MatejLach @grainloom @Shamar @cbowdon

I think it's important to make it clear what we mean by committee.

When I say committee, I think one where each person tries to achieve their own goals at all costs, and push the whole design into a direction that's more favorable to them.

If you have people honestly cooperate and be open to logical arguments, pointing out each others mistakes and willing to be proven wrong, that's another story.

@a_breakin_glass @Wolf480pl @z428 @MatejLach @grainloom @alcinnz @cbowdon

What about BOFH? 🤣

Just kidding...

The point of what we are saying is that if everyone in the planet was able to hack his operating system (or whatever), you woudn't need BDFL. You woudn't need maintainers or funding. And no corporation could win against free software.

@Shamar @Wolf480pl @z428 @MatejLach @grainloom @alcinnz @cbowdon
>You woudn't need maintainers or funding.

you might; after all, you need infrastructure to share patches, etc over

@a_breakin_glass @Wolf480pl @z428 @MatejLach @grainloom @alcinnz @cbowdon

In a world where each house have more servers that rooms? I guess we won't need much more.

@Shamar @a_breakin_glass @Wolf480pl @z428 @MatejLach @alcinnz @cbowdon This works best if the modifications they make don't mess things up. Eg. if I have a patched version of ls but other software wants to use ls, that might be a problem. So we need good package management tools.

@Wolf480pl @Shamar @a_breakin_glass @z428 @MatejLach @alcinnz @cbowdon I still haven't been able to try NixOS because the VM kept running out memory but uuuh, it's probably what I want? Or at least close.

@Wolf480pl @cbowdon @alcinnz @MatejLach @z428 @a_breakin_glass @Shamar @grainloom

what i think we need is a well-defined OS Protocol Suite and a component-oriented system architecture. ideally you'd have a very small and simple core that can host a larger set of components that provide all of the services and functionality that you expect from a modern OS.

Genode explores this concept using L4-style IPC. the design is honestly very appealing, but C++ isn't something i'm interested in using.
@Wolf480pl @grainloom @Shamar @a_breakin_glass @z428 @MatejLach @alcinnz @cbowdon

something Genode-compatible in C

or an adaptation of Plan 9 to a microkernel architecture

HURD isn't very interesting to me tbh

@xj9 @grainloom @Shamar @a_breakin_glass @z428 @MatejLach @alcinnz @cbowdon
I read some Hurd's design docs and parts of code some time ago, and came to the conclusion that it is basically an adaptation of Plan9 to a microkernel.

@xj9 @Wolf480pl @cbowdon @alcinnz @MatejLach @z428 @a_breakin_glass @grainloom

I looked for that source code recently but wasn't able to find it anywhere. Jehanne is going to be very similar to what you describe in the long term. Tho microkernel or not, it won't be much relevant on a distributed system as many kernels can coexist: the operating system is the network.

@Shamar @grainloom @a_breakin_glass @z428 @MatejLach @alcinnz @cbowdon @Wolf480pl

i'm into microkernel here for security reasons. you don't need it to participate on the network os, but idk why you'd want to underpin your network system with a less a than secure legacy kernel.

@xj9 @grainloom @a_breakin_glass @z428 @MatejLach @alcinnz @cbowdon @Wolf480pl

Uhm... I agree on this but to be honest I didn't found a microkernel with enough advantages to justify a total rewrite of Jehanne's one.

I really welcome suggestions.

Constrains are LP64, little endian with SMP support written in C or a simpler language.

@Shamar @Wolf480pl @cbowdon @alcinnz @MatejLach @z428 @a_breakin_glass @grainloom

that’s chill. it’s better if there are many implementations of the standard. whatever that ends up being. extensions to 9p or some other protocol might be a good place to start.

@xj9 @Wolf480pl @cbowdon @alcinnz @MatejLach @z428 @a_breakin_glass @grainloom

Uhm... I don't know what "chill" means in this context.

Multiple implementation will be welcome (it's a totally new protocol) but it's not going to become a standard: that would open to non-copylefted implementations and I want the next Web to be Free Software at every single layer.

@Shamar @grainloom @a_breakin_glass @z428 @MatejLach @alcinnz @cbowdon @Wolf480pl

so you want a viral protocol license? i guess that would work. i suppose we wont be able to interop then since my projects are all Unlicense.

oh well ¯\_(ツ)_/¯
Show more
Show more

@Shamar @xj9 @cbowdon @alcinnz @MatejLach @z428 @a_breakin_glass @grainloom

IANAL, but from what I understand, under current copyright law network protocols temselves are not copyrightable. I can always get one person to reverse-engineer your program and write a spec for the protocol in such an uncreative way that the spec is not copyrightable, and then have another person read the spec and make a clean-room impl.
And there are copyright exceptions for "reverse engineering for compatibility".

Show more
Show more

@a_breakin_glass @z428 @MatejLach @grainloom @Shamar @alcinnz @cbowdon

It is at least internally consistent.

The result of design-by-committee usually ends up being a bunch of mutually incompatible ideas bolted together with duct tape. See: SQL.

Or something way more complicated than it needs to. See X.500.

@Wolf480pl @MatejLach @grainloom @z428 @alcinnz @cbowdon


This would disrupt programming as a job?

No, we still have novelists, journalists and secretaries despite everybody is able to write.

Would it disrupt the Silicon Valley.

Yes. Yes it would. 😇

@Wolf480pl That software they use on a daily basis also includes things such as power management, device drivers or firmware. Where do you draw the line? What about that "bug" of your BIOS always throttling your CPU "too much" when you're off AC?
@MatejLach @grainloom @Shamar @alcinnz @cbowdon

@z428 @Wolf480pl @MatejLach @grainloom @Shamar @cbowdon I'd say wherever they are comfortable drawing the line.

We shouldn't be the ones to draw it. But until some of these breakthroughs Shamer's talking about happens at the hardware level, their line almost certainly won't include device drivers.

@alcinnz That's what I mean. And also possibly not network protocols or stuff such as multithreading. It seems strange to want to make solutions to complex problems randomly easy. Why can't ordinary users do the math to build a highrise that doesn't collapse? Because it's complex. As a problem. Not just because we lack better tools.
@Wolf480pl @MatejLach @grainloom @Shamar @cbowdon

@z428 @MatejLach @grainloom @Shamar @alcinnz @cbowdon
The only line I draw is between "it does X in this particular case, let's see if we can make it do Y instead" and "this code is pretty complicated, let's see if we can come up with a better design that simplifies this part and everything that touches it".

I don't draw the line between a messaging app and power management code in the kernel.

@Wolf480pl Yes. That happens regularly, even in the programming world done by experts, and it fails almost all the time because people tried to "simplify" an inherently complex thing they just considered too complex because they never managed to fully understand even the problem it tried to solve. See CORBA vs. SOAP. ;)
@MatejLach @grainloom @Shamar @alcinnz @cbowdon

@z428 @MatejLach @grainloom @Shamar @alcinnz @cbowdon

I think this is why, when contributing to a FOSS project, or doing ad-hoc modifications to locally-installed versions of programs you use, you usually start with bugfixes, small tweaks, maybe some small features.

I think it'd be cool if more people got to that level.
And I think it's ok if most people stay at that level.
Not everyone needs to be able to refactor things or rewrite them from scratch.

@Wolf480pl @z428 @MatejLach @grainloom @alcinnz @cbowdon

Yeah... I think most people should be content they have been teached to write their name and to modify a shopping list.

What do they want?
Writing open letters like Émile Zola?
By themselves?

It's unwise... they would end in trouble because, let's be honest, they are not smart enough!
It's better they trust us, the Writing class, to think for them.

@phryk @z428

I didn't intend to say that @Wolf480pl is elitist, the whole IT is because our field is so primitive to require decades to be fluent.

But we shouldn't settle on this, even just from an egoistic perspective.

If we cannot explain what we know simply, we don't know it well enough. Our elitism is a monument to our collective #ignorance. But despite it we don't understand or accept our power and responsibility.

We could change the world through #Informatics.

@Shamar Yes..... and *this* is the actual problem: We don't want inclusive and enabling technology that gives a lot of users abilities that are easily accessible. We want them to learn our way (knowing they never will even remotely be able to walk it) rather than using our "superiority" to watch them, listen to them and help them solve problems. That ...

@Wolf480pl @MatejLach @grainloom @alcinnz @cbowdon

@z428 @Wolf480pl @MatejLach @grainloom @alcinnz @cbowdon

Changing master won't make you #free.

Not even if the new one is (temporarily) a benevolent class of nerds like us.

Talking about #FreeSoftware but assuming that you need a degree to read and modify it has a name: #hypocrisy.


@z428 @Wolf480pl @MatejLach @grainloom @alcinnz @cbowdon

You are right that in the short term being kind and welcoming nerds is all we can do. We should also care about older people that are never going to understand how they are strictly controlled and used by the #software they use.

But if we don't change approach, this will never change. And who benefit from such #power?
A bunch of #BrainWashing corps.

@Shamar The essence of what you are saying is: Specialize in everything so you can do everything on your own to be free and do not have to depend upon anyone because no one can be trusted (why should this be limited to software?). Good luck trying.
@Wolf480pl @MatejLach @grainloom @alcinnz @cbowdon

@z428 @Wolf480pl @MatejLach @grainloom @alcinnz @cbowdon

I'm saying that programming is not going to be different from writing: it's something that enhance our ability to reason, that literally augment the abilities of our minds and that insisting to reserve it to a cast of elect is not just elitism but plain shortsighted.

It's like we were trying to reserve Math to a cast because "they can't get it".

@Shamar That's a pretty much different statement compared to what you wrote earlier and *miles* away from people being able to understand or even fix the software they use on a daily basis.
@Wolf480pl @MatejLach @grainloom @alcinnz @cbowdon

@z428 @Wolf480pl @MatejLach @grainloom @alcinnz @cbowdon

To me it is exactly the same.

The difference between explaining De Morgan laws to a class and make them deduce it by themselves while writing a program is that in one case most students will forget it after the exam, in the others most students will consider them obvious and integrate them in their way of thinking.

If you can't do, you don't know.

@Shamar No. A lot of our users in planning and construction use software such as AutoCAD for designing buildings all day. The best they possibly could do is some superficial scripted automation. They *never* would be able to fix anything in this application, even if they had the sources. Where is your point?
@Wolf480pl @MatejLach @grainloom @alcinnz @cbowdon

Sign in to participate in the conversation
Matej Lach's mastodon

The social network of the future: No ads, no corporate surveillance, ethical design, and decentralization! Own your data with Mastodon!